Imagine a summer weekend on a California beach dense with bodies. But for one onlooker, this seemingly calm scene may be a series of accidents waiting to happen. How does a lifeguard know when a raised arm means "I need help", not "hey,this is fun"? The guard's skill at spotting that one desperate person among thousands is phenomenal, truly testing his sight and understanding.
We see the way we do, so we can spot danger to ourselves. But nothing is directing the lifeguard. In fact the eye, observing a harmless pattern across its view, normally relaxes. Motion-sensing rod cells switch off when they detect action that's consistent and constant. So the lifeguard has to trick his eyes. He does this by scanning, forcing his eyes to lock onto small details.
Our frontline defence are the tower guards, and their job is to scan the water. So their eyes are moving across the water and letting their brains filter out that information that they see, looking for something wrong, looking for that odd one out that truly is in danger.
Taking in all this information is hard work. Human sight has only two degrees of detail vision at the center. To check the whole beach, the lifeguard sweeps, jumping from point to point for detail. Each jump is called a saccade.
A saccade is the movement that the eyes make together when they're looking directly at one thing, and all of a sudden, they look at something else. We have mechanisms that wire the muscles that move our eyes to the image, and we can quickly lock onto a new image all at once.
The saccade function lets him jump visually from each potential risk to the next. He repeatedly scans his field of vision, updating his visual memory every few seconds.